Seeing Bird Sounds

How looking at spectrograms can help us hear more

Dan Brown

GGAS Master Birder Student Salon

2019-11-15

Recordings from xeno-canto.org used under Creative Commons license

These slides available at jdanbrown.org/talks/ggas-mb

What is a spectrogram?

• Light to dark — loudness
• Left to right — time
• Low to high — pitch

_spectro_scale = 5
_selected_recs = {
    243973: ('GCSP', 'descending, high-pitched whistles'),
    172196: ('GHOW', 'low-pitched hoots'),
    154721: ('FOSP', 'short chips, variable tempo'),
    408021: ('WREN', 'short "chirps", increasing tempo'),
    252944: ('STJA', 'harsh calls (many frequencies)'),
    110976: ('CASJ', 'harsh calls, upward pitch'),
    188039: ('WCSP', 'whistle → burry → downslurs'),
    # 174545: ('SOSP', '"tink" notes → fast trill → burry whistle'),
}
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x | y, [
        *[
            (df.species == species) & (df.xc_id == xc_id)
            for xc_id, (species, _) in _selected_recs.items()
        ],
        # df.species == 'WREN',  # XXX For explore
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    # .sample(frac=1, random_state=0)  # XXX For explore
    .assign(_sortby=lambda df: df.xc_id.pipe(lambda s: as_ordered_cat(s, _selected_recs.keys())))
    .sort_values('_sortby')
    # .sort_values('xc_id')  # XXX For explore
    # View
    .assign(_talk_notes=lambda df: df.xc_id.map(lambda xc_id: '' if xc_id not in _selected_recs else _selected_recs[xc_id][-1]))
    .pipe(talk_recs_show,
        scale=_spectro_scale,
        append=['_talk_notes'],
    )
    [:100]
    .pipe(add_recs_to_cite)
    [[
        # 'xc_id', 'species',  # XXX For explore
        'species_com_name',
        'spectro_disp',
        '_talk_notes',
    ]]

    # The normal, simple thing
    # .pipe(talk_hide_index_and_columns)

    # HACK Un-kill td padding for not spectros
    # HACK Crop spectro img widths (doesn't change audio)
    .pipe(lambda df: display_with_style(
        talk_hide_index_and_columns(df),
        style_css=lambda scoped_class: f'''
            /* HACK Un-kill td padding for not spectros */
            .{scoped_class} .dataframe tbody tr td:nth-child(2),
            .{scoped_class} .dataframe tbody tr td:nth-child(4) {{
                padding: 0.2em 0.5em 0.2em 0.5em !important;
            }}
            /* HACK Crop spectro img widths (doesn't change audio) */
            .{scoped_class} .bubo-audio-container {{
                width: {4/10 * _spectro_scale * spectro_img_width_px}px;
                overflow: hidden;
            }}
        ''',
    ))

)

(52613, 71)

(7, 71)

	species_com_name	spectro_disp	_talk_notes
0	Golden-crowned Sparrow		descending, high-pitched whistles
1	Great Horned Owl		low-pitched hoots
2	Fox Sparrow		short chips, variable tempo
3	Wrentit		short "chirps", increasing tempo
4	Steller's Jay		harsh calls (many frequencies)
5	California Scrub-Jay		harsh calls, upward pitch
6	White-crowned Sparrow		whistle → burry → downslurs

What does a Lesser Goldfinch sound like?

What does a Lesser Goldfinch sound like?

• What do we see?

_spectro_scale = 5
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x & y, [
        df.species == 'LEGO',
        df.quality.isin(['A', 'B']),
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .sample(frac=1, random_state=2)
    # View
    .pipe(talk_recs_show,
        scale=_spectro_scale,
    )
    [:50]
    .pipe(add_recs_to_cite)
    [['spectro_disp']]
    .pipe(talk_hide_index_and_columns)
)

(52613, 71)

(110, 71)

	spectro_disp
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

Let's learn a few birds

Pacific Slope Flycacher

• What do we see?

_spectro_scale = 5
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x & y, [
        df.species == 'PSFL',
        df.quality.isin(['A']),
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .sample(frac=1, random_state=0)
    # View
    .pipe(talk_recs_show,
        scale=_spectro_scale,
    )
    [:50]
    .pipe(add_recs_to_cite)
    [['spectro_disp']]
    .pipe(talk_hide_index_and_columns)
)

(52613, 71)

(67, 71)

	spectro_disp
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

Hutton's Vireo

• What do we see?

_spectro_scale = 5
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x & y, [
        df.species == 'HUVI',
        df.quality.isin(['A', 'B']),
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .sample(frac=1, random_state=2)
    # View
    .pipe(talk_recs_show,
        scale=_spectro_scale,
    )
    [:50]  # (148 recs in quality=A,B)
    .pipe(add_recs_to_cite)
    [['spectro_disp']]
    .pipe(talk_hide_index_and_columns)
)

(52613, 71)

(148, 71)

	spectro_disp
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

Bewick's Wren (a harder example)

• What do we see?

_spectro_scale = 5
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x & y, [
        df.species == 'BEWR',
        df.quality.isin(['A']),
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .sample(frac=1, random_state=1)
    # View
    .pipe(talk_recs_show,
        scale=_spectro_scale,
    )
    [:50]
    .pipe(add_recs_to_cite)
    [['spectro_disp']]
    .pipe(talk_hide_index_and_columns)
)

(52613, 71)

(168, 71)

	spectro_disp
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

Let's dig a little deeper

Spotted Towhee: “west coast” vs. “interior”

• What do we see?

_spectro_scale = 4
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x & y, [
        df.species == 'SPTO',
        # df.type.str.lower().str.match(r'.*song.*'),
        # df.type.str.lower().str.match(r'^song$'),
        df.quality.isin(['A', 'B']),
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    # .sample(250, random_state=0)
    # .sample(50, random_state=0)
    # .sort_values(['subspecies', 'lat'], ascending=False)
    # .sort_values(['lat'], ascending=False)
    # .sort_values(['lng'], ascending=True)
    .pipe(df_inspect, lambda df: (df.shape,))
    # View
    .pipe(talk_recs_show,
        scale=_spectro_scale,
    )

    # Bin by geo (2 bins)
    #   - Very rough, but decent results
    .assign(geo=lambda df: df.pipe(df_map_rows, lambda row: (
        'West of -117°: no intro notes → fast/slow trill' if row.lng < -117 else  # -117 to exclude Inyo County (Death Valley) which behaves ~like interior
        'East of -110°: intro notes → slow trill'         if row.lng > -110 else  # -110 to include CO/AZ/MX (at least)
        None                                                                      # Throw out everything in the middle, to simplify analysis
    )))
    [lambda df: df.geo.notnull()]
    .assign(geo=lambda df: df.geo.pipe(lambda s: s.pipe(as_ordered_cat, [
        'West of -117°: no intro notes → fast/slow trill',
        'East of -110°: intro notes → slow trill',
    ])))
    # .pipe(df_inspect, lambda df: df[:250])  # Debug

    # Pivot by geo (manually)
    .pipe(lambda df: pd.concat(axis=1, objs=[
        (df
            [df.geo == geo]
            .rename(columns={'spectro_disp': geo})
            .sample(frac=1, random_state=0)  # Randomize per column (to avoid weird biases from incidental sorting above)
            .reset_index(drop=True)
            [:25]
            .pipe(add_recs_to_cite)
            [[geo]]
        )
        for geo in df.geo.sort_values().unique()
    ]))
    .fillna('')
    .pipe(df_inspect, lambda df: (df.shape,))

    # The normal, simple thing
    # .pipe(talk_hide_index)

    # HACK Crop spectro img widths (doesn't change audio)
    .pipe(lambda df: display_with_style(
        talk_hide_index(df),
        style_css=lambda scoped_class: f'''
            /* HACK Crop spectro img widths (doesn't change audio) */
            .{scoped_class} .bubo-audio-container {{
                width: {6/10 * _spectro_scale * spectro_img_width_px}px;
                overflow: hidden;
            }}
        ''',
    ))

)

(52613, 71)

(432, 71)

(432, 71)

(25, 2)

	West of -117°: no intro notes → fast/slow trill	East of -110°: intro notes → slow trill
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

What goes “wraa”?

# 1
_spectro_scale = 4
_cmap_name = (
    'Pastel1'
    # 'Pastel2'
    # 'Set3'
)
_selected_recs = {
    14843:  ('SPTO', '"wraa" / mew'),
    127012: ('SPTO', '"wraa" / mew'),
}
_colors = [mpl.colors.rgb2hex(x) for x in getattr(mpl.cm, _cmap_name).colors]
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x | y, [
        *[
            (df.species == species) & (df.xc_id == xc_id)
            for xc_id, (species, _) in _selected_recs.items()
        ],
        # df.species == 'HUVI',  # XXX For explore
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .assign(_sortby=lambda df: df.xc_id.pipe(lambda s: as_ordered_cat(s, _selected_recs.keys())))
    .sort_values('_sortby')
    # .sort_values('xc_id')  # XXX For explore
    # .sample(frac=1, random_state=0)  # XXX For explore

    # View
    .assign(_talk_notes=lambda df: df.xc_id.map(lambda xc_id: '' if xc_id not in _selected_recs else _selected_recs[xc_id][-1]))
    .pipe(talk_recs_show,
        scale=_spectro_scale,
        append=['_talk_notes'],
    )
    [:100]
    .pipe(add_recs_to_cite)
    [[
        # 'xc_id', 'species',  # XXX For explore
        'species_com_name',
        '_talk_notes',
        'spectro_disp',
    ]]

    # The normal, simple thing
    # .pipe(talk_hide_index_and_columns)

    # HACK Un-kill td padding for not spectros
    # HACK Color rows by species
    .assign(_color=lambda df: df.species_com_name.pipe(lambda s: (
        s.map({x: _colors[i] for i, x in enumerate(unique_everseen(s.tolist()))}.__getitem__)
    )))
    .pipe(lambda df: display_with_style(
        talk_hide_index_and_columns(df
            .drop(columns=['_color'])
        ),
        style_css=lambda scoped_class: '\n'.join([
            # HACK Restrict background to first two cols (after index col)
            f'''
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(2),
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(3) {{
                    padding: 0.2em 0.5em 0.2em 0.5em !important; /* HACK Un-kill td padding for not spectros */
                    background: {_color} !important; /* HACK Color rows by species *//
                }}
            '''
            for i, _color in enumerate(df._color.tolist())
        ]),
    ))

)

(52613, 71)

(2, 71)

	species_com_name	_talk_notes	spectro_disp
0	Spotted Towhee	"wraa" / mew
1	Spotted Towhee	"wraa" / mew

What goes “wraa”?

# 2
_spectro_scale = 4
_cmap_name = (
    'Pastel1'
    # 'Pastel2'
    # 'Set3'
)
_selected_recs = {
    14843:  ('SPTO', '"wraa" / mew'),
    127012: ('SPTO', '"wraa" / mew'),
    297119: ('HUVI', '"wraa"'),
    297120: ('HUVI', '"wraa"'),
    # 348986: ('HUVI', '"wraa"'),
}
_colors = [mpl.colors.rgb2hex(x) for x in getattr(mpl.cm, _cmap_name).colors]
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x | y, [
        *[
            (df.species == species) & (df.xc_id == xc_id)
            for xc_id, (species, _) in _selected_recs.items()
        ],
        # df.species == 'HUVI',  # XXX For explore
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .assign(_sortby=lambda df: df.xc_id.pipe(lambda s: as_ordered_cat(s, _selected_recs.keys())))
    .sort_values('_sortby')
    # .sort_values('xc_id')  # XXX For explore
    # .sample(frac=1, random_state=0)  # XXX For explore

    # View
    .assign(_talk_notes=lambda df: df.xc_id.map(lambda xc_id: '' if xc_id not in _selected_recs else _selected_recs[xc_id][-1]))
    .pipe(talk_recs_show,
        scale=_spectro_scale,
        append=['_talk_notes'],
    )
    [:100]
    .pipe(add_recs_to_cite)
    [[
        # 'xc_id', 'species',  # XXX For explore
        'species_com_name',
        '_talk_notes',
        'spectro_disp',
    ]]

    # The normal, simple thing
    # .pipe(talk_hide_index_and_columns)

    # HACK Un-kill td padding for not spectros
    # HACK Color rows by species
    .assign(_color=lambda df: df.species_com_name.pipe(lambda s: (
        s.map({x: _colors[i] for i, x in enumerate(unique_everseen(s.tolist()))}.__getitem__)
    )))
    .pipe(lambda df: display_with_style(
        talk_hide_index_and_columns(df
            .drop(columns=['_color'])
        ),
        style_css=lambda scoped_class: '\n'.join([
            # HACK Restrict background to first two cols (after index col)
            f'''
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(2),
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(3) {{
                    padding: 0.2em 0.5em 0.2em 0.5em !important; /* HACK Un-kill td padding for not spectros */
                    background: {_color} !important; /* HACK Color rows by species *//
                }}
            '''
            for i, _color in enumerate(df._color.tolist())
        ]),
    ))

)

(52613, 71)

(4, 71)

	species_com_name	_talk_notes	spectro_disp
0	Spotted Towhee	"wraa" / mew
1	Spotted Towhee	"wraa" / mew
2	Hutton's Vireo	"wraa"
3	Hutton's Vireo	"wraa"

What goes “wraa”?

# 3
_spectro_scale = 4
_cmap_name = (
    'Pastel1'
    # 'Pastel2'
    # 'Set3'
)
_selected_recs = {
    14843:  ('SPTO', '"wraa" / mew'),
    127012: ('SPTO', '"wraa" / mew'),
    297119: ('HUVI', '"wraa"'),
    297120: ('HUVI', '"wraa"'),
    # 348986: ('HUVI', '"wraa"'),
    268597: ('HETH', 'hoarse wheeze'),
    13677:  ('HETH', 'hoarse wheeze'),
    # 131636: ('HETH', 'hoarse wheeze'),
}
_colors = [mpl.colors.rgb2hex(x) for x in getattr(mpl.cm, _cmap_name).colors]
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x | y, [
        *[
            (df.species == species) & (df.xc_id == xc_id)
            for xc_id, (species, _) in _selected_recs.items()
        ],
        # df.species == 'HUVI',  # XXX For explore
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .assign(_sortby=lambda df: df.xc_id.pipe(lambda s: as_ordered_cat(s, _selected_recs.keys())))
    .sort_values('_sortby')
    # .sort_values('xc_id')  # XXX For explore
    # .sample(frac=1, random_state=0)  # XXX For explore

    # View
    .assign(_talk_notes=lambda df: df.xc_id.map(lambda xc_id: '' if xc_id not in _selected_recs else _selected_recs[xc_id][-1]))
    .pipe(talk_recs_show,
        scale=_spectro_scale,
        append=['_talk_notes'],
    )
    [:100]
    .pipe(add_recs_to_cite)
    [[
        # 'xc_id', 'species',  # XXX For explore
        'species_com_name',
        '_talk_notes',
        'spectro_disp',
    ]]

    # The normal, simple thing
    # .pipe(talk_hide_index_and_columns)

    # HACK Un-kill td padding for not spectros
    # HACK Color rows by species
    .assign(_color=lambda df: df.species_com_name.pipe(lambda s: (
        s.map({x: _colors[i] for i, x in enumerate(unique_everseen(s.tolist()))}.__getitem__)
    )))
    .pipe(lambda df: display_with_style(
        talk_hide_index_and_columns(df
            .drop(columns=['_color'])
        ),
        style_css=lambda scoped_class: '\n'.join([
            # HACK Restrict background to first two cols (after index col)
            f'''
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(2),
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(3) {{
                    padding: 0.2em 0.5em 0.2em 0.5em !important; /* HACK Un-kill td padding for not spectros */
                    background: {_color} !important; /* HACK Color rows by species *//
                }}
            '''
            for i, _color in enumerate(df._color.tolist())
        ]),
    ))

)

(52613, 71)

(6, 71)

	species_com_name	_talk_notes	spectro_disp
0	Spotted Towhee	"wraa" / mew
1	Spotted Towhee	"wraa" / mew
2	Hutton's Vireo	"wraa"
3	Hutton's Vireo	"wraa"
4	Hermit Thrush	hoarse wheeze
5	Hermit Thrush	hoarse wheeze

What goes “wraa”?

# 4
_spectro_scale = 4
_cmap_name = (
    'Pastel1'
    # 'Pastel2'
    # 'Set3'
)
_selected_recs = {
    14843:  ('SPTO', '"wraa" / mew'),
    127012: ('SPTO', '"wraa" / mew'),
    297119: ('HUVI', '"wraa"'),
    297120: ('HUVI', '"wraa"'),
    # 348986: ('HUVI', '"wraa"'),
    268597: ('HETH', 'hoarse wheeze'),
    13677:  ('HETH', 'hoarse wheeze'),
    # 131636: ('HETH', 'hoarse wheeze'),
    159366: ('WAVI', '"wraa"'),
    381527: ('WAVI', '"wraa"'),
    141349: ('BEWR', 'harsh buzz'),
    163209: ('BEWR', 'harsh buzz'),
    265810: ('HOWR', 'harsh buzz #2'),
    376229: ('BGGN', 'polyphonic wheeze'),
    81059:  ('BGGN', 'polyphonic wheeze'),
}
_colors = [mpl.colors.rgb2hex(x) for x in getattr(mpl.cm, _cmap_name).colors]
(search_recs
    .pipe(df_inspect, lambda df: (df.shape,))
    # Filter
    [lambda df: reduce(lambda x, y: x | y, [
        *[
            (df.species == species) & (df.xc_id == xc_id)
            for xc_id, (species, _) in _selected_recs.items()
        ],
        # df.species == 'HUVI',  # XXX For explore
    ])]
    .pipe(df_inspect, lambda df: (df.shape,))
    # Sort
    .assign(_sortby=lambda df: df.xc_id.pipe(lambda s: as_ordered_cat(s, _selected_recs.keys())))
    .sort_values('_sortby')
    # .sort_values('xc_id')  # XXX For explore
    # .sample(frac=1, random_state=0)  # XXX For explore

    # View
    .assign(_talk_notes=lambda df: df.xc_id.map(lambda xc_id: '' if xc_id not in _selected_recs else _selected_recs[xc_id][-1]))
    .pipe(talk_recs_show,
        scale=_spectro_scale,
        append=['_talk_notes'],
    )
    [:100]
    .pipe(add_recs_to_cite)
    [[
        # 'xc_id', 'species',  # XXX For explore
        'species_com_name',
        '_talk_notes',
        'spectro_disp',
    ]]

    # The normal, simple thing
    # .pipe(talk_hide_index_and_columns)

    # HACK Un-kill td padding for not spectros
    # HACK Color rows by species
    .assign(_color=lambda df: df.species_com_name.pipe(lambda s: (
        s.map({x: _colors[i] for i, x in enumerate(unique_everseen(s.tolist()))}.__getitem__)
    )))
    .pipe(lambda df: display_with_style(
        talk_hide_index_and_columns(df
            .drop(columns=['_color'])
        ),
        style_css=lambda scoped_class: '\n'.join([
            # HACK Restrict background to first two cols (after index col)
            f'''
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(2),
                .{scoped_class} tbody tr:nth-child({i + 1}) td:nth-child(3) {{
                    padding: 0.2em 0.5em 0.2em 0.5em !important; /* HACK Un-kill td padding for not spectros */
                    background: {_color} !important; /* HACK Color rows by species *//
                }}
            '''
            for i, _color in enumerate(df._color.tolist())
        ]),
    ))

)

(52613, 71)

(13, 71)

	species_com_name	_talk_notes	spectro_disp
0	Spotted Towhee	"wraa" / mew
1	Spotted Towhee	"wraa" / mew
2	Hutton's Vireo	"wraa"
3	Hutton's Vireo	"wraa"
4	Hermit Thrush	hoarse wheeze
5	Hermit Thrush	hoarse wheeze
6	Warbling Vireo	"wraa"
7	Warbling Vireo	"wraa"
8	Bewick's Wren	harsh buzz
9	Bewick's Wren	harsh buzz
10	House Wren	harsh buzz #2
11	Blue-gray Gnatcatcher	polyphonic wheeze
12	Blue-gray Gnatcatcher	polyphonic wheeze

Thanks!

• Questions?

# Credits
HTML('''
    <div class="talk-credits">
        %(children)s
    </div>
''' % dict(
    children=(recs_to_cite
        .pipe(df_map_rows, lambda row: '''
            <div class="talk-credits-credit">
                <a href="https://www.xeno-canto.org/%(xc_id)s">%(species_com_name)s: XC%(xc_id)s © %(recordist)s</a>
            </div>
        ''' % row)
        .pipe(lambda s: '\n'.join(s.tolist()))
    ),
))

Golden-crowned Sparrow: XC243973 © Davyd Betchkal

Great Horned Owl: XC172196 © Micah Riegner

Fox Sparrow: XC154721 © Mike Schanbacher

Wrentit: XC408021 © Frank Lambert

Steller's Jay: XC252944 © Manuel Grosselet

California Scrub-Jay: XC110976 © Jonathon Jongsma

White-crowned Sparrow: XC188039 © Ted Floyd

Lesser Goldfinch: XC219941 © Paul Marvin

Lesser Goldfinch: XC219939 © Paul Marvin

Lesser Goldfinch: XC219953 © Paul Marvin

Lesser Goldfinch: XC220450 © Peter Boesman

Lesser Goldfinch: XC141493 © Mike Nelson

Lesser Goldfinch: XC181098 © Elvis Herrera Rodríguez

Lesser Goldfinch: XC254541 © Manuel Grosselet

Lesser Goldfinch: XC147186 © Garrett MacDonald

Lesser Goldfinch: XC220454 © Peter Boesman

Lesser Goldfinch: XC254490 © Nick Komar

Lesser Goldfinch: XC94655 © Scott Olmstead

Lesser Goldfinch: XC324919 © Bruce Lagerquist

Lesser Goldfinch: XC71956 © Richard E Webster

Lesser Goldfinch: XC301686 © Nick Komar

Lesser Goldfinch: XC219950 © Paul Marvin

Lesser Goldfinch: XC219951 © Paul Marvin

Lesser Goldfinch: XC141492 © Mike Nelson

Lesser Goldfinch: XC220455 © Peter Boesman

Lesser Goldfinch: XC219946 © Paul Marvin

Lesser Goldfinch: XC219947 © Paul Marvin

Lesser Goldfinch: XC219943 © Paul Marvin

Lesser Goldfinch: XC279568 © Nick Komar

Lesser Goldfinch: XC220456 © Peter Boesman

Lesser Goldfinch: XC73808 © Ryan P. O'Donnell

Lesser Goldfinch: XC71959 © Richard E Webster

Lesser Goldfinch: XC219938 © Paul Marvin

Lesser Goldfinch: XC297493 © Ross Gallardy

Lesser Goldfinch: XC219933 © Paul Marvin

Lesser Goldfinch: XC331908 © Esau Toaki Villarreal

Lesser Goldfinch: XC219952 © Paul Marvin

Lesser Goldfinch: XC352735 © Frank Lambert

Lesser Goldfinch: XC146581 © Manuel Grosselet

Lesser Goldfinch: XC71957 © Richard E Webster

Lesser Goldfinch: XC133956 © Lauren Harter

Lesser Goldfinch: XC331817 © Sue Riffe

Lesser Goldfinch: XC229823 © Peter Boesman

Lesser Goldfinch: XC14548 © Andrew Spencer

Lesser Goldfinch: XC279566 © Nick Komar

Lesser Goldfinch: XC220445 © Peter Boesman

Lesser Goldfinch: XC269512 © Richard E. Webster

Lesser Goldfinch: XC102954 © Eric DeFonso

Lesser Goldfinch: XC375747 © Bobby Wilcox

Lesser Goldfinch: XC219940 © Paul Marvin

Lesser Goldfinch: XC146611 © Dan Lane

Lesser Goldfinch: XC121404 © Lauren Harter

Lesser Goldfinch: XC71953 © Richard E Webster

Lesser Goldfinch: XC374080 © Matt Wistrand

Lesser Goldfinch: XC219934 © Paul Marvin

Lesser Goldfinch: XC219929 © Paul Marvin

Lesser Goldfinch: XC348358 © Richard E. Webster

Pacific-slope Flycatcher: XC104715 © Paul van Els

Pacific-slope Flycatcher: XC71886 © Richard E Webster

Pacific-slope Flycatcher: XC123997 © Richard E. Webster

Pacific-slope Flycatcher: XC348144 © Richard E. Webster

Pacific-slope Flycatcher: XC348150 © Richard E. Webster

Pacific-slope Flycatcher: XC71885 © Richard E Webster

Pacific-slope Flycatcher: XC124960 © Richard E. Webster

Pacific-slope Flycatcher: XC71881 © Richard E Webster

Pacific-slope Flycatcher: XC71877 © Richard E Webster

Pacific-slope Flycatcher: XC297476 © Ross Gallardy

Pacific-slope Flycatcher: XC71897 © Richard E Webster

Pacific-slope Flycatcher: XC71873 © Richard E Webster

Pacific-slope Flycatcher: XC71882 © Richard E Webster

Pacific-slope Flycatcher: XC71871 © Richard E Webster

Pacific-slope Flycatcher: XC400450 © Richard E. Webster

Pacific-slope Flycatcher: XC71878 © Richard E Webster

Pacific-slope Flycatcher: XC71893 © Richard E Webster

Pacific-slope Flycatcher: XC400447 © Richard E. Webster

Pacific-slope Flycatcher: XC156767 © Ian Cruickshank

Pacific-slope Flycatcher: XC71889 © Richard E Webster

Pacific-slope Flycatcher: XC37051 © Tayler Brooks

Pacific-slope Flycatcher: XC348149 © Richard E. Webster

Pacific-slope Flycatcher: XC353028 © Richard E. Webster

Pacific-slope Flycatcher: XC71883 © Richard E Webster

Pacific-slope Flycatcher: XC348142 © Richard E. Webster

Pacific-slope Flycatcher: XC408438 © Frank Lambert

Pacific-slope Flycatcher: XC142826 © Ian Cruickshank

Pacific-slope Flycatcher: XC348139 © Richard E. Webster

Pacific-slope Flycatcher: XC348151 © Richard E. Webster

Pacific-slope Flycatcher: XC348153 © Richard E. Webster

Pacific-slope Flycatcher: XC71892 © Richard E Webster

Pacific-slope Flycatcher: XC353035 © Richard E. Webster

Pacific-slope Flycatcher: XC71884 © Richard E Webster

Pacific-slope Flycatcher: XC124972 © Richard E. Webster

Pacific-slope Flycatcher: XC178530 © Thomas G. Graves

Pacific-slope Flycatcher: XC182545 © Ian Cruickshank

Pacific-slope Flycatcher: XC408490 © Frank Lambert

Pacific-slope Flycatcher: XC386589 © Skip Russell

Pacific-slope Flycatcher: XC124969 © Richard E. Webster

Pacific-slope Flycatcher: XC71895 © Richard E Webster

Pacific-slope Flycatcher: XC408437 © Frank Lambert

Pacific-slope Flycatcher: XC65992 © Ian Cruickshank

Pacific-slope Flycatcher: XC297477 © Ross Gallardy

Pacific-slope Flycatcher: XC34289 © Tayler Brooks

Pacific-slope Flycatcher: XC124964 © Richard E. Webster

Pacific-slope Flycatcher: XC71896 © Richard E Webster

Pacific-slope Flycatcher: XC71872 © Richard E Webster

Pacific-slope Flycatcher: XC71890 © Richard E Webster

Pacific-slope Flycatcher: XC159100 © Ian Cruickshank

Pacific-slope Flycatcher: XC178942 © Eric Cannizzaro

Hutton's Vireo: XC135962 © Dan Lane

Hutton's Vireo: XC232135 © Peter Boesman

Hutton's Vireo: XC282857 © Eric Hough

Hutton's Vireo: XC71052 © Mary Beth Stowe

Hutton's Vireo: XC171209 © Micah Riegner

Hutton's Vireo: XC266858 © Manuel Grosselet

Hutton's Vireo: XC305574 © James Bradley

Hutton's Vireo: XC254339 © Randy Dzenkiw

Hutton's Vireo: XC348986 © Paul Marvin

Hutton's Vireo: XC352238 © Frank Lambert

Hutton's Vireo: XC283520 © David Vander Pluym

Hutton's Vireo: XC232138 © Peter Boesman

Hutton's Vireo: XC348987 © Paul Marvin

Hutton's Vireo: XC168038 © Manuel Grosselet

Hutton's Vireo: XC279002 © Nick Komar

Hutton's Vireo: XC1230 © Don Jones

Hutton's Vireo: XC196558 © Dan Lane

Hutton's Vireo: XC21394 © Richard E Webster

Hutton's Vireo: XC306671 © Jarrod Swackhamer

Hutton's Vireo: XC29358 © Andrew Spencer

Hutton's Vireo: XC55202 © manuel Grosselet

Hutton's Vireo: XC323414 © Richard E. Webster

Hutton's Vireo: XC113920 © Lance A. M. Benner

Hutton's Vireo: XC232144 © Peter Boesman

Hutton's Vireo: XC282597 © David Vander Pluym

Hutton's Vireo: XC70993 © Mary Beth Stowe

Hutton's Vireo: XC235009 © Garrett MacDonald

Hutton's Vireo: XC323411 © Richard E. Webster

Hutton's Vireo: XC105702 © Nathan Pieplow

Hutton's Vireo: XC231587 © Peter Boesman

Hutton's Vireo: XC323407 © Richard E. Webster

Hutton's Vireo: XC158510 © Ian Cruickshank

Hutton's Vireo: XC406527 © Thomas G. Graves

Hutton's Vireo: XC232146 © Peter Boesman

Hutton's Vireo: XC21397 © Richard E Webster

Hutton's Vireo: XC417129 © Manuel Grosselet

Hutton's Vireo: XC189330 © Ian Cruickshank

Hutton's Vireo: XC21395 © Richard E Webster

Hutton's Vireo: XC296970 © Paul Marvin

Hutton's Vireo: XC30595 © Andrew Spencer

Hutton's Vireo: XC159564 © Ian Cruickshank

Hutton's Vireo: XC253906 © Bobby Wilcox

Hutton's Vireo: XC34324 © Andrew Spencer

Hutton's Vireo: XC183585 © Ian Cruickshank

Hutton's Vireo: XC136113 © Dan Lane

Hutton's Vireo: XC282598 © David Vander Pluym

Hutton's Vireo: XC181337 © Tim Marquardt

Hutton's Vireo: XC327652 © Joshua Jaeger

Hutton's Vireo: XC372121 © markbird

Hutton's Vireo: XC156794 © Ian Cruickshank

Bewick's Wren: XC159932 © Ian Cruickshank

Bewick's Wren: XC171094 © Elvis Herrera Rodríguez

Bewick's Wren: XC254552 © Manuel Grosselet

Bewick's Wren: XC351081 © Richard E. Webster

Bewick's Wren: XC217835 © Paul Marvin

Bewick's Wren: XC173153 © Eric DeFonso

Bewick's Wren: XC383782 © Manuel Grosselet

Bewick's Wren: XC41230 © Scott Olmstead

Bewick's Wren: XC159353 © Ian Cruickshank

Bewick's Wren: XC205433 © Eric DeFonso

Bewick's Wren: XC22069 © Chris Parrish

Bewick's Wren: XC321497 © Joshua Jaeger

Bewick's Wren: XC323979 © Richard E. Webster

Bewick's Wren: XC217824 © Paul Marvin

Bewick's Wren: XC212943 © Jim Holmes

Bewick's Wren: XC163210 © Garrett MacDonald

Bewick's Wren: XC371801 © Manuel Grosselet

Bewick's Wren: XC164259 © Ian Cruickshank

Bewick's Wren: XC158936 © Ian Cruickshank

Bewick's Wren: XC172752 © Eric DeFonso

Bewick's Wren: XC104638 © Andrew Spencer

Bewick's Wren: XC277908 © Eric Hough

Bewick's Wren: XC274880 © Frank Lambert

Bewick's Wren: XC183590 © Ian Cruickshank

Bewick's Wren: XC178015 © Thomas G. Graves

Bewick's Wren: XC351087 © Richard E. Webster

Bewick's Wren: XC296660 © Laura Gooch

Bewick's Wren: XC254537 © Manuel Grosselet

Bewick's Wren: XC370889 © Russ Wigh

Bewick's Wren: XC160282 © Ian Cruickshank

Bewick's Wren: XC408559 © Frank Lambert

Bewick's Wren: XC323975 © Richard E. Webster

Bewick's Wren: XC123370 © Chris Harrison

Bewick's Wren: XC159988 © Ian Cruickshank

Bewick's Wren: XC217858 © Paul Marvin

Bewick's Wren: XC360348 © Jim Holmes

Bewick's Wren: XC168584 © Manuel Grosselet

Bewick's Wren: XC351086 © Richard E. Webster

Bewick's Wren: XC122453 © Chris Harrison

Bewick's Wren: XC323981 © Richard E. Webster

Bewick's Wren: XC159811 © Ian Cruickshank

Bewick's Wren: XC12577 © Andrew Spencer

Bewick's Wren: XC230953 © Peter Boesman

Bewick's Wren: XC160250 © Ian Cruickshank

Bewick's Wren: XC351058 © Richard E. Webster

Bewick's Wren: XC390104 © Guy Kirwan

Bewick's Wren: XC217816 © Paul Marvin

Bewick's Wren: XC217819 © Paul Marvin

Bewick's Wren: XC22068 © Chris Parrish

Bewick's Wren: XC173154 © Eric DeFonso

Spotted Towhee: XC14843 © Andrew Spencer

Spotted Towhee: XC127012 © Richard E. Webster

Hutton's Vireo: XC297119 © Paul Marvin

Hutton's Vireo: XC297120 © Paul Marvin

Hermit Thrush: XC268597 © Frank Lambert

Hermit Thrush: XC13677 © Andrew Spencer

Warbling Vireo: XC159366 © Ian Cruickshank

Warbling Vireo: XC381527 © Bobby Wilcox

Bewick's Wren: XC141349 © Mike Nelson

Bewick's Wren: XC163209 © Garrett MacDonald

House Wren: XC265810 © Andrew Spencer

Blue-gray Gnatcatcher: XC376229 © Ted Floyd

Blue-gray Gnatcatcher: XC81059 © Ryan P. O'Donnell

Spotted Towhee: XC381708 © Jeremy Minns

Spotted Towhee: XC158943 © Ian Cruickshank

Spotted Towhee: XC160245 © Ian Cruickshank

Spotted Towhee: XC305616 © Nick Komar

Spotted Towhee: XC73501 © Ian Cruickshank

Spotted Towhee: XC126990 © Richard E. Webster

Spotted Towhee: XC413996 © Thomas G. Graves

Spotted Towhee: XC127003 © Richard E. Webster

Spotted Towhee: XC100251 © Trent R. Stanley

Spotted Towhee: XC159417 © Ian Cruickshank

Spotted Towhee: XC76885 © Stuart Fisher

Spotted Towhee: XC159879 © Ian Cruickshank

Spotted Towhee: XC126665 © Richard E. Webster

Spotted Towhee: XC408095 © Frank Lambert

Spotted Towhee: XC124488 © Thomas G. Graves

Spotted Towhee: XC381707 © Jeremy Minns

Spotted Towhee: XC142824 © Ian Cruickshank

Spotted Towhee: XC179116 © Ian Cruickshank

Spotted Towhee: XC126694 © Richard E. Webster

Spotted Towhee: XC159552 © Ian Cruickshank

Spotted Towhee: XC325228 © James Bradley

Spotted Towhee: XC160135 © Ian Cruickshank

Spotted Towhee: XC195846 © Elias Aristides Elias

Spotted Towhee: XC364820 © Bruce Lagerquist

Spotted Towhee: XC159909 © Ian Cruickshank

Spotted Towhee: XC112070 © Richard E Webster

Spotted Towhee: XC367803 © Eric DeFonso

Spotted Towhee: XC65642 © Daniel Lane

Spotted Towhee: XC368589 © Sue Riffe

Spotted Towhee: XC112082 © Richard E Webster

Spotted Towhee: XC111826 © Richard E Webster

Spotted Towhee: XC252813 © Manuel Grosselet

Spotted Towhee: XC12585 © Andrew Spencer

Spotted Towhee: XC323292 © Richard E. Webster

Spotted Towhee: XC205751 © Eric DeFonso

Spotted Towhee: XC103315 © Eric DeFonso

Spotted Towhee: XC228178 © Peter Boesman

Spotted Towhee: XC389884 © Leonardo Guzman Hernandez

Spotted Towhee: XC376227 © Ted Floyd

Spotted Towhee: XC14426 © Andrew Spencer

Spotted Towhee: XC248103 © Manuel Grosselet

Spotted Towhee: XC112115 © Richard E Webster

Spotted Towhee: XC323289 © Richard E. Webster

Spotted Towhee: XC168046 © Manuel Grosselet

Spotted Towhee: XC32580 © manuel Grosselet

Spotted Towhee: XC205280 © Eric DeFonso

Spotted Towhee: XC112137 © Richard E Webster

Spotted Towhee: XC348671 © Matt Wistrand

Spotted Towhee: XC70184 © manuel Grosselet

Spotted Towhee: XC112054 © Richard E Webster